Abstract In this paper we develop a many-valued and multi-preferential conditional logic with typicality, based on a multi-preferential semantics, which generalizes KLM preferential semantics. We then exploit the multi-preferential semantics to provide preferential interpretation of gradual argumentation. The approach allows for conditional reasoning over arguments and boolean combination of arguments, with respect to some chosen gradual semantics, through the verification of graded (strict or defeasible) implications over an argumentation graph. The paper also develops a probabilistic semantics for gradual argumentation, which builds on the many-valued semantics.

In this paper we investigate the relationships between a multipreferential semantics for defeasible reasoning in knowledge representation and a multilayer neural network model. Weighted knowledge bases for a simple description logic with typicality are considered under a (many-valued) “concept-wise” multipreference semantics. The semantics is used to provide a preferential interpretation of MultiLayer Perceptrons (MLPs). A model checking and an entailment based approach are exploited in the verification of conditional properties of MLPs.

Abstract Weighted knowledge bases for description logics with typicality under a ‘concept-wise’ multi-preferential semantics provide a logical interpretation of MultiLayer Perceptrons. In this context, Answer Set Programming (ASP) has been shown to be suitable for addressing defeasible reasoning in the finitely many-valued case, providing a $\varPi ^{p}_{2}$ upper bound on the complexity of the problem, nonetheless leaving unknown the exact complexity and only providing a proof-of-concept implementation. This paper fulfills the lack by providing a ${P^{NP[log]}}$-completeness result and new ASP encodings that deal with both acyclic and cyclic weighted knowledge bases with large search spaces, as assessed empirically on synthetic test cases. The encodings are used to empower a reasoner for computing solutions and answering queries, possibly interacting with ASP Chef for obtaining an interactive visualization.

The ANGELIC methodology was successfully used to predict decisions of the European Court of Human Rights based on a set of logical rules, with significantly better accuracy than the one achieved by machine learning approaches, as well as to explain the results of reasoning, quite valuable in order to make them trustworthy. This work demonstrates a different logic-based approach, based on Answer Set Programming for solving and generating explanations for solutions. The use of a general knowledge representation and reasoning system, where representation and inference are not tightly coupled, allows for using the same representation for inference tasks different from prediction, thus getting more value out of the domain model, and opens for integrating further forms of knowledge.

Abstract In this paper we establish a link between fuzzy and preferential semantics for description logics and self-organizing maps (SOMs), which have been proposed as possible candidates to explain the psychological mechanisms underlying category generalization. In particular, we show that the input/output behavior of a SOM after training can be described by a fuzzy description logic interpretation as well as by a preferential interpretation, based on a concept-wise multipreference semantics, which takes into account preferences with respect to different concepts and has been recently proposed for ranked and for weighted defeasible description logics. Properties of the network can be proven by model checking on the fuzzy or on the preferential interpretation. Starting from the fuzzy interpretation, we also provide a probabilistic account for this neural network model.

Abstract Weighted knowledge bases for description logics with typicality have been recently considered under a “concept-wise” multipreference semantics (in both the two-valued and fuzzy case), as the basis of a logical semantics of multilayer perceptrons (MLPs). In this paper we consider weighted conditional $\mathcal{ALC}$ knowledge bases with typicality in the finitely many-valued case, through three different semantic constructions. For the boolean fragment $\mathcal{LC}$ of $\mathcal{ALC}$ we exploit answer set programming and asprin for reasoning with the concept-wise multipreference entailment under a $\varphi$ -coherent semantics, suitable to characterize the stationary states of MLPs. As a proof of concept, we experiment the proposed approach for checking properties of trained MLPs.

We propose a framework for structuring the description and results of the forensic analysis of actions of investigative interest in digital applications, and for automated reasoning on such actions. A high level of abstraction is suitable for forensic stakeholders that are not ICT experts; other levels are suitable for automating experiments on the devices to establish traces left by actions, and for associating the results of the experiments. Such results are used in a computational logic framework to conclude evidence on the occurrence of actions. The evidence can be presented to stakeholders or used in further automated reasoning, and traced back to data on the device.

Abstract In this paper we develop a concept aware multi-preferential semantics for dealing with typicality in description logics, where preferences are associated with concepts, starting from a collection of ranked TBoxes containing defeasible concept inclusions. Preferences are combined to define a preferential interpretation in which defeasible inclusions can be evaluated. The construction of the concept-aware multipreference semantics is related to Brewka’s framework for qualitative preferences. We exploit Answer Set Programming (in particular,asprin) to achieve defeasible reasoning under the multipreference approach for the lightweight description logic ξ $\mathcal L_ \bot ^ + $ .

In this work we study a rational extension 𝒮ℛ𝒪ℰℒ(⊓, ×)R T of the low complexity description logic 𝒮ℛ𝒪ℰℒ(⊓, ×), which underlies the OWL EL ontology language. The extension involves a typicality operator T, whose semantics is based on Lehmann and Magidor’s ranked models and allows for the definition of defeasible inclusions. We consider both rational entailment and minimal entailment. We show that deciding instance checking under minimal entailment is in general ∏2P -hard, while, under rational entailment, instance checking can be computed in polynomial time. We develop a Datalog calculus for instance checking under rational entailment and exploit it, with stratified negation, for computing the rational closure of simple KBs in polynomial time.

Clinical Guidelines (CGs) provide general evidence-based recommendations and physicians often have to resort also to their Basic Medical Knowledge (BMK) to cope with specific patients. In this paper, we explore the interplay between CGs and BMK from the viewpoint of a-posteriori conformance analysis, intended as the adherence of a specific execution log to both the CG and the BMK. In this paper, we consider also the temporal dimension: the guideline may include temporal constraints for the execution of actions, and its adaptation to a specific patient and context may add or modify conditions and temporal constraints for actions. We propose an approach for analyzing execution traces in Answer Set Programming with respect to a guideline and BMK, pointing out discrepancies - including temporal discrepancies - with respect to the different knowledge sources, and providing explanations regarding how the applications of the CG and the BMK have interacted, especially in case strictly adhering to both is not possible.

Abstract In this paper we exploit Answer Set Programming (ASP) for reasoning in a rational extensionSROEL(⊓,×)RTof the low complexity description logicSROEL(⊓, ×), which underlies the OWL EL ontology language. In the extended language, a typicality operatorTis allowed to define conceptsT(C) (typicalC's) under a rational semantics. It has been proven that instance checking under rational entailment has a polynomial complexity. To strengthen rational entailment, in this paper we consider a minimal model semantics. We show that, for arbitrarySROEL(⊓,×)RTknowledge bases, instance checking under minimal entailment is ΠP2-complete. Relying on a Small Model result, where models correspond to answer sets of a suitable ASP encoding, we exploit Answer Set Preferences (and, in particular, theasprinframework) for reasoning under minimal entailment.

Temporal logics are well suited for reasoning about actions, as they allow for the specification of domain descriptions including temporal constraints as well as for the verification of temporal properties. The article deals with verification of action theories defined in a temporal extension of answer set programming which combines ASP with a dynamic linear time temporal logic (DLTL). The article proposes an approach to bounded model checking that exploits the Büchi automaton construction while searching for a counterexample, with the aim of achieving completeness. The article provides an encoding in ASP of the temporal action domains and of Bounded Model Checking of DLTL formulas. The article also deals with reasoning about epistemic knowledge and incomplete states.

Computer Interpretable Guidelines (CIG) are an emerging area of research, to support medical decision making through evidence-based recommendations. However, new challenges in the data management field have to be faced, to integrate CIG management with a proper treatment of patient data, and of other forms of medical knowledge (e.g., causal and behavioral knowledge). In this position paper, we summarize a proposal for a research agenda that, in our opinion, can lead to a significant advancement in the field. The goal of the work is to provide suitable models and reasoning methodologies to cope with the aforementioned aspects, and to properly integrate them for medical decision support. Achieving such a goal requires advances in data management, and, in particular, in the treatment of indeterminate valid-time data in relational databases, of temporal abstraction on time series, of case retrieval on time series, of design-time and run-time model-based verification of guidelines, of case-based reasoning, of non-monotonic logics, of formal ontologies, of probabilistic graphical models (Bayesian Networks and Influence Diagrams).

The verification of compliance of business processes to norms requires the representation of different kinds of obligations, including achievement obligations, maintenance obligations, obligations with deadlines and contrary to duty obligations. In this paper we develop a deontic temporal extension of Answer Set Programming (ASP) suitable for verifying compliance of a business process to norms involving such different types of obligations. To this end, we extend Dynamic Linear Time Temporal Logic (DLTL) with deontic modalities to define a Deontic DLTL. We then combine it with ASP to define a deontic action language in which until formulas and next formulas are allowed to occur within deontic modalities. We show that in the language we can model the different kinds of obligations which are useful in the verification of compliance to normative requirements. The verification can be performed by bounded model checking techniques.

This article introduces a framework for creating rich augmented environments based on a social web of intelligent things and people. We target outdoor environments, aiming to transform a region into a smart environment that can share its cultural heritage with people, promoting itself and its special qualities. Using the applications developed in the framework, people can interact with things, listen to the stories that these things tell them, and make their own contributions. The things are intelligent in the sense that they aggregate information provided by users and behave in a socially active way. They can autonomously establish social relationships on the basis of their properties and their interaction with users. Hence when a user gets in touch with a thing, she is also introduced to its social network consisting of other things and of users; she can navigate this network to discover and explore the world around the thing itself. Thus the system supports serendipitous navigation in a network of things and people that evolves according to the behavior of users. An innovative interaction model was defined that allows users to interact with objects in a natural, playful way using smartphones without the need for a specially created infrastructure. The framework was instantiated into a suite of applications called WantEat, in which objects from the domain of tourism and gastronomy (such as cheese wheels or bottles of wine) are taken as testimonials of the cultural roots of a region. WantEat includes an application that allows the definition and registration of things, a mobile application that allows users to interact with things, and an application that supports stakeholders in getting feedback about the things that they have registered in the system. WantEat was developed and tested in a real-world context which involved a region and gastronomy-related items from it (such as products, shops, restaurants, and recipes), through an early evaluation with stakeholders and a final evaluation with hundreds of users.

Abstract In this paper, we combine Answer Set Programming (ASP) with Dynamic Linear Time Temporal Logic (DLTL) to define a temporal logic programming language for reasoning about complex actions and infinite computations. DLTL extends propositional temporal logic of linear time with regular programs of propositional dynamic logic, which are used for indexing temporal modalities. The action language allows general DLTL formulas to be included in domain descriptions to constrain the space of possible extensions. We introduce a notion of Temporal Answer Set for domain descriptions, based on the usual notion of Answer Set. Also, we provide a translation of domain descriptions into standard ASP and use Bounded Model Checking (BMC) techniques for the verification of DLTL constraints.

Abstract The paper provides a framework for the verification of business processes, based on an extension of answer set programming (ASP) with temporal logic and constraints. The framework allows to capture expressive fluent annotations as well as data awareness in a uniform way. It allows for a declarative specification of a business process but also for encoding processes specified in conventional workflow languages. Verification of temporal properties of a business process, including verification of compliance to business rules, is performed by bounded model checking techniques in Answer Set Programming, extended with constraint solving for dealing with conditions on numeric data.

Wanteat is a framework and a suite of applications which allow users to interact with and explore mixed social networks of smart objects and people in the gastronomy domain, thus promoting the cultural heritage of a territory. Wanteat interaction model is based on the concept of a "wheel" [1].

Several explanation and interpretation tasks, such as diagnosis, plan recognition and image interpretation, can be formalized as abductive reasoning. A number of approaches, including recent ones [1, 4], address the problem based on a task-independent representation of a domain which includes an ontology or taxonomy of hypotheses. In this paper we adopt a similar representation, but we also deal with abduction as an iterative process where, like in model-based diagnosis, further observations are proposed to discriminate among candidate explanations; in addition, we take into account costs of observations and actions. In fact, discrimination also involves refining hypotheses, but this is performed down to an appropriate level which depends on the cost of actions (e.g. repair actions or therapy) to be taken based on the results of abduction, and on the cost of additional observations, which should be balanced with the benefits, in terms of more suitable actions, of better discrimination. The presence of a domain representation with abstractions has a significant impact on this trade-off. In general, a better assessment of the situation at hand, based on additional observations, leads to a more focused action. However, the cost of observing the same phenomenon at different levels of abstraction may vary significantly; in fact, it could involve more or less costly medical or technical tests, or computationally complex image processing, possibly with additional costs due to the delay before taking an action. Moreover, the knowledge base could have been designed independently of the explanation/action task (e.g. diagnosis and repair), and could therefore include a detailed description of the domain which is not necessary for the task; more generally, the convenience of a detailed discrimination may depend on the specific case at hand. By explicitly considering abstractions in the iterative abduction process, we can often reduce the observation costs significantly, yet maintaining the ability to exploit detailed observations and knowledge when convenient (similar advantages have been shown in inductive classification with abstractions, e.g. [6]). In the following, we first describe the knowledge we expect to be available. We then describe a basic iterative abduction loop and, finally, we concentrate on the criterion for selecting the next step in the loop: either performing a next observation at some level of detail, or stopping because the estimated most convenient choice is performing the action(s) associated with the current hypotheses.

The automatic generation of decision trees based on off-line reasoning on models of a domain is a reasonable compromise between the advantages of using a model-based approach in technical domains and the constraints imposed by embedded applications. In this paper we extend the approach to deal with temporal information. We introduce a notion of temporal decision tree, which is designed to make use of relevant information as long as it is acquired, and we present an algorithm for compiling such trees from a model-based reasoning system.

Generating abductive explanations is the basis of several problem solving activities such as diagnosis, planning, and interpretation. Temporal abduction means generating explanations that do not only account for the presence of observations, but also for temporal information on them, based on temporal knowledge in the domain theory. We focus on the case where such a theory contains temporal constraints that are required to be consistent with temporal information on observations. Our aim is to propose efficient algorithms for computing temporal abductive explanations. Temporal constraints in the theory and in the observations can be used actively by an abductive reasoner in order to prune inconsistent candidate explanations at an early stage during their generation. However, checking temporal constraint satisfaction frequently generates some overhead. We analyze two incremental ways of making this process efficient. First we show how, using a specific class of temporal constraints (which is expressive enough for many applications), such an overhead can be reduced significantly, yet preserving a full pruning power. In general, the approach does not affect the asymptotic complexity of the problem, but it provides significant advantages in practical cases. We also show that, for some special classes of theories, the asymptotic complexity is also reduced. We then show how, compiled knowledge based on temporal information, can be used to further improve the computation, thus, extending to the temporal framework previous results in the case of atemporal abduction. The paper provides both analytic and experimental evaluations of the computational advantages provided by our approaches.

Several artificial intelligence architectures and systems based on "deep" models of a domain have been proposed, in particular for the diagnostic task. These systems have several advantages over traditional knowledge based systems, but they have a main limitation in their computational complexity. One of the ways to face this problem is to rely on a knowledge compilation phase, which produces knowledge that can be used more effectively with respect to the original one. We show how a specific knowledge compilation approach can focus reasoning in abductive diagnosis, and, in particular, can improve the performances of AID, an abductive diagnosis system. The approach aims at focusing the overall diagnostic cycle in two interdependent ways: avoiding the generation of candidate solutions to be discarded a posteriori and integrating the generation of candidate solutions with discrimination among different candidates. Knowledge compilation is used off-line to produce operational (i.e., easily evaluated) conditions that embed the abductive reasoning strategy and are used in addition to the original model, with the goal of ruling out parts of the search space or focusing on parts of it. The conditions are useful to solve most cases using less time for computing the same solutions, yet preserving all the power of the model-based system for dealing with multiple faults and explaining the solutions. Experimental results showing the advantages of the approach are presented.

The aim of this paper is to analyse from various points of view the relationships between abduction and deduction. In particular, we consider a meta-level definition of abduction in terms of deduction, similar to various definitions proposed in the literature, and an object-level definition in which abductive conclusions are expressed as a logicalconsequence of the observations and of a simple transformation of the domain theory based on predicate completion. The equivalence between the two definitions is proved for domain theories of considerable expressive power. The object-level characterization we propose uses very simple forms of reasoning and the equivalence result allows us to make explicit some of the assumptions underlying meta-level definitions of abduction. The use of predicate completion in characterizing abductive explanations shows a relation between abduction and foundations of logic programming.